Given concerns about the appropriate diagnosis for children with chronic, severe irritability, we defined the syndrome of severe mood dysregulation(SMD)to capture youth with extremely severe irritability and symptoms of hyperarousal. Our SMD phenotype formed the basis for the new pediatric diagnosis of mood dysregulation disorder with dysphoria(DMDD)in DSM-5. Since the inception of this project, approximately 300 youth with SMD/DMDD have been recruited into the project. Approximately 20 new patients were recruited this year. Youth with DMDD suffer very severe psychiatric impairment, as assessed by medications received, number of psychiatric hospitalizations, and standardized measures of function. As noted above, we demonstrated differences between youth with SMD and those with bipolar disorder (BD) in clinical course, family history, and brain mechanisms associated with symptoms. While we continue to compare youth with BD and those with severe irritability in terms of mediating brain mechanisms, we have become increasingly interested in the pathophysiology of irritability in its own right. Indeed, irritability is one of the most common presenting symptoms in pediatric psychiatry. The construct of irritability is particularly well-suited for the transdiagnostic, translational approach of the Research Domain Criteria (RDoC). Much of our work now involves examining irritability, not only in DMDD, but also in other groups exhibiting irritability, including youth with anxiety disorders, ADHD, or BD. Across groups, we characterize irritability as a continuous variable. A focus of our neuroimaging work includes the use of frustrating tasks, since a hallmark of irritability is difficulty tolerating frustration. Previously, we demonstrated behavioral and neural differences between irritable and non-irritable youth during a frustrating attentional task. We improved that paradigm and obtained fMRI data from approximately 200 youth with DMDD, BD, ADHD, and/or anxiety disorders, as well as healthy youth; these data are currently being analyzed. This approach allows us to examine the neural correlates of irritability, operationalized dimensionally, while also comparing DSM-5 diagnostic groups in the brain circuitry engaged during a frustrating task. We also demonstrated in a community-based sample (N=109) that the same frustrating task has good test-retest reliability in inducing frustration (therefore, it might be used pre- and post-treatment as a measure of treatment response). Also, the child's ratings of his/her frustration during the task correlated with his/her ratings of irritability in daily life, supporting the suggestion that the task assesses processes relevant to the child's symptoms. We have also piloted an additional frustration task both inside and outside the scanner, to try to improve even further our ability to detect neural mechanisms mediating frustration. We continue our work comparing the neural circuitry mediating face emotion processing among youth with DMDD, BD, and healthy youth. As with the frustration work, we now incorporate a dimensional approach, examining neural associations with continuous measures of irritability. In work published this year, we compared the impact of irritability on brain activation across youth with BD, DMDD, and healthy subjects. Brain activity was measured while youth performed a labeling task with happy, fearful, and angry faces of varying emotional intensity. We found differences between BD and DMDD in associations between irritability and brain activity. Specifically, irritability correlated with amygdala activity across all intensities for all emotions in the DMDD group; such correlation was present in the BD group only for fearful faces. In the ventral visual stream, associations between neural activity and irritability were found more consistently in the DMDD group than in the BD group, especially in response to ambiguous angry faces. These data suggest that the circuitry mediating a dimension, such as irritability, might vary across diagnoses. RDoC does not explicitly incorporate a longitudinal perspective; arguably, such a perspective is more central to the diagnosis of BD than to other diagnoses, such as DMDD, anxiety disorders, and ADHD. Therefore, compared to the latter three diagnoses, cross-sectional dimensions alone may not be less able to explain the circuitry of a trait when that trait appears within the context of BD. Second, these data suggest that treatment approaches, such as one described below designed to modify irritable children's responses to angry faces, might be less effective in irritable youth with BD than in those with DMDD. This work emphasizes the importance of integrating both categorical (i.e., diagnostic group) and dimensional (e.g., irritability) research approaches. Another unexplored aspect of the RDoC framework is whether traits interact neurally. We are interested in possible interactions between anxiety and irritability, given cross-sectional and longitudinal associations between these traits and possible shared circuitry mediating them. In healthy youth and those with DMDD, ADHD, anxiety disorders (N=115), we found that irritability and anxiety jointly influenced left amygdala to left medial prefrontal cortex (mPFC) connectivity during viewing of intensely angry faces. Specifically, decreased connectivity was associated with high levels of both anxiety and irritability, whereas increased connectivity was associated with high levels of anxiety but low levels of irritability. This study yields the important conclusion that the circuitry mediating dimensional traits should not be considered in isolation from each other. A major focus of our work is treatment. We continue our double-blind trial designed to ascertain whether citalopram (a serotonergic reuptake inhibitor (SRI) antidepressant that is effective in the treatment of pediatric anxiety) plus stimulant is more effective than placebo plus stimulant in the treatment of severe irritability. Stimulant and SRI treatment tend to have fewer side-effects than atypical antipsychotic treatment, which that is considered first-line treatment for BD, yet stimulants and SRI's are relatively contraindicated in patients with BD because of concern about possibly inducing a manic episode. Therefore, this work has considerable public health importance. We have randomized approximately 50 children into the trial, and youth are tolerating the treatment well. We are now developing and testing two novel psychosocial approaches to treating irritability. The first involves computer-based training designed to shift a subject's perception of ambiguous faces from angry toward happy. In work published this year, we found that youth with DMDD tend to judge ambiguous faces as angry rather than happy; that we can change this judgment with computer-based training; and that this is associated with decreased irritability. This was an open trial, without a control condition. We have begun a controlled trial of this new Interpretation Bias Treatment, along with brain imaging before and after training. Recruitment has been robust. We are also now piloting a manual-based cognitive behavioral treatment targeting irritability. This has a strong emphasis on behavioral techniques i.e., exposing the child to tolerable, but frustrating situations, to improve the child's ability to tolerate such situations. The treatment is based on our neuroscience-based model of irritability, including irritable children's hypothesized deficits in the process and content of instrumental learning. We are now piloting a neuroimaging paradigm designed to assess such deficits in irritable youth. Such a paradigm could that can be used pre- and post- treatment to document changes in the circuitry mediating these processes during successful treatment.